gfprep
[-nqvdpPBU] [-mx] [-X regexp
] [-S source-domainname
] [-h source-hostfile
] [-D destination-domainname
] [-H destination-hostfile
] [-j num-of-processes
] [-J num-of-processes
] [-M total-byte
] [-z minimum-byte
] [-Z maximum-byte
] [-w way-of-scheduling
] [-W kilobytes-for-threshold
] [-s KB/s-to-simulate
] [-F num-for-readahead
] [-I sec-to-update
] [-N num-of-replicas
] path
gfprep creates replicas of Gfarm files in parallel.
When the path
specifies a
directory, files under the directory will be replicated recursively.
Path
can be specified by a
Gfarm URL or a path on a mount point of gfarm2fs.
A set of source/destination hosts can be specified by a domain name and/or a hostlist file. When both a domain name and a hostlist file are specified, a set of hosts is determined by both conditions. When a set of source hosts is specified, only files stored on the source hosts are replicated.
Less busy file system nodes will be selected for source or destination. The maximum number of parallel replications/copies per host is the number of CPUs. Please see the manual page of gfhost for information about the number of CPUs.
gfprep also retrieves the directory entries in parallel.
These are options only for gfprep.
-N
num-of-replicas
Specifies the required number of file replicas. If there are a sufficient number of replicas in the destination hosts, neither any copying nor any removal will be done.
If this option is not specified, -N 1 is assumed.
-x
Removes extra file replicas when a set of destination nodes has more file replicas than specified by the -N option.
-m
Migrates file replicas from a set of source nodes to destination nodes. When specifying twice like -mm, it leaves the source file replica.
This option requires at least one of -S, -h, -D, and -H options.
This option and the -N option are mutually exclusive.
The following options are common options for gfprep and gfpcopy.
-X
regexp
Skips files matched by the pattern
of regexp
. When multiple
patterns need to be specified, specify -X options multiple times.
-S
source-domainname
Creates file replicas or copies files only stored on the hosts in the specified domain name.
-h
source-hostfile
Creates file replicas or copies files only stored on the hosts listed
in the specified hostfile.
The source-hostfile
consists of
a file system node name on each line.
If ``-'' is specified, standard input is used to read the host list.
-L
Creates file replicas or copies files from the hosts specified by the -S or -h option.
-D
destination-domainname
Specifies the domain name for destination.
If neither this nor the -H
option is specified,
replicas may be copied to any available host.
-H
destination-hostfile
Specifies a file which describes hostnames for destination.
The destination-hostfile
consists of a file system node name on each line.
If ``-'' is specified, standard input is used to read the host list.
-j
num-of-processes
Specifies the maximum number of processes to create file replicas (or copy files) simultaneously.
The default value is the parameter of client_parallel_copy in gfarm2.conf. (see man gfarm2.conf)
The maximum number of process per file system node for source or destination is the number of CPUs (see man gfhost).
-J
num-of-processes
Specifies the number of processes to retrieve directory entries in parallel.
The default value is 8.
-M
total-byte
Specifies the total file size in bytes to replicate or copy. This option is useful to increase the available capacity by moving the specified bytes of files.
The default value is unlimited.
-z
minimum-byte
Specifies the minimum file size in bytes to replicate or copy. This option is useful not to replicate or copy small files.
The default value is unlimited.
-Z
maximum-byte
Specifies the maximum file size in bytes to replicate or copy. This option is useful not to replicate or copy large files.
The default value is unlimited.
-w
way-of-scheduling
Specifies a scheduling method. ``noplan'' replicates/copies while finding files. ``greedy'' schedules greedily the order of replication/copy beforehand.
The default behavior is ``noplan''.
``greedy'' scheduling cannot use with the -N option and -m option.
-W
kibibytes
Specifies a threshold size/cost(KiB) to flat costs of Connections. A Connection means a scheduling information to assign files per a child-process
This option is effective with -w greedy.
The default value is 50*1024 KiB (50 MiB).
-I
sec-to-update
Specifies the interval in seconds to collect load average and available capacity.
Default is 300 seconds.
-B
Gfarm 2.6.16 or later does not select high loaded file system nodes. This option disables this feature.
High loaded node is defined by having more CPU load than schedule_busy_load_thresh * number of CPUs. For details of schedule_busy_load_thresh, refer to a manual page of gfarm2.conf.
-U
Disables checking the available disk space of the selected node every time.
-F
num-of-dirents
Specifies the number of readahead entries to retrieve the directory entries.
The default value is 10000.
-s
kilobytes-per-second
Specifies a throughput(KB/s) to simulate the replication/copy, and does nothing (gets file information only).
-n
Does nothing.
-p
Reports the total performance information.
-P
Reports the performance information for each file and all files.
-q
Suppresses non-error messages.
-v
Displays verbose output.
-d
Displays debug output.
-?
Displays a list of command options.
To replicate files under the directory recursively.
$ gfprep -N 3 gfarm:///dir
$ cd /mnt/gfarm2fs
$ gfprep -N 3 dir
To replicate a file or remove surplus replicas of the file.
$ gfprep -N 3 -x gfarm:///dir/file
To migrate replicas from gfsd1.example.com to other nodes.
$ gfprep -m -S gfsd1.example.com gfarm:///
When the -L or -m option is not specified, a source node is not always selected within a set of specified source file system nodes.
It is better to execute gfprep command near the metadata server. When you need to execute gfprep command far from the metadata server, increase the parallelism by the -j and -J options.